This will be quantitative study of how the retinal image in cats is sampled by three classes of ganglion cells. These cells (X, Y and Q cells) form the vast majority of cat ganglion cells with center-surround receptive fields, and, for X and Y at least, are a major component of the retinogeniculostriate pathway. An important aspect of the proposed study is that the spatiotemporal filetering properties will be determined as a function of retinal position, so that the retinal inhomogeneity in filtering properties can be quantified. In the same retinas that these filtering properties of cells are measured, the sampling arrays of the X and the Y cells from the points where the recordings were made will be measured anatomically, allowing comparison between retinal inhomogeneities in sampling and in filtering properties. With these data, a complete map of how the retinal image is sampled by these cells will be obtained. This map should be of considerable use in modeling the spatiotemporal properties of receptive fields of neurons located at higher levels of the visual system and ultimately in modeling spatial vision, binocular vision and stereopsis. It will also help more immdiately in quantifying descriptions of local visual field loss due to retinal damage resulting from diseases of the eye (e.g gaucoma, maculopathies).